Predicting out-terminals for imported containers at seaports using machine learning - Incorporating unstructured data and measuring operational costs due to misclassifications

Aug 4, 2026

Upload by Quinn

Persistent bottlenecks at container ports disrupt global supply chains, making seaport operations more efficient while managing yard density and congestion. An approach is to use container characteristics with machine learning to predict a container’s out-terminal after discharge. The prediction can support better container storage strategy. The research proposes a data-driven framework with structured and unstructured data integration, knowledge-informed feature engineering, explanatory classification models, and cost analysis of misclassifications, validated with 14.90%-30.45% savings, improved performance up to 6%, and low operational integration risk.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages23

Reading Time58 min

AccessFree to download

Introduction

Motivation

Proposed methodological framework

Data integration and feature engineering

Explanatory machine learning classification models

Cost analysis for misclassification impacts

Empirical study and results

Cost savings and performance improvement

Operational integration feasibility

Broader implications

Transportation Research Part E 202 (2025) 104331 Contents lists available at ScienceDirect Transportation Research Part E journal [homepage: www.elsevier.com/locate/tre](homepage: www.elsevier.com/locate/tre) | Predicting out-terminals for imported containers at seaports using machine learning: Incorporating unstructured data and measuring operational costs due to misclassifications
Ying Xiea, Dong-Ping Song b,*, Jingxin Dongc, Yuanjun Feng b
a Cranfield School of Management, Cranfield University, UK b School of Management, University of Liverpool, UK c Business School, Newcastle University, UK | | | | --- | --- | --- | | A R T I C L E I N F O | | A B S T R A C T | | Keywords:
Seaport
Container classification Predictive model
Feature engineering Explanatory machine learning Cost analysis | | Persistent bottlenecks at container ports have significantly disrupted global supply chains, necessitating more efficient operations at seaports to address yard density and port congestion. An untapped but potentially critical approach to mitigating these challenges is to leverage container characteristics and machine learning to predict the out-terminals of containers upon their discharge from vessels. The predicted results can then guide the development of a more effective container storage strategy. To formulate such a strategy, this research developed a data-enabled methodological framework that integrates four key components: 1) Utilization of structured and unstructured data to enhance prediction accuracy. 2) Practice and knowledge-informed feature engineering to construct relevant features for the machine learning models. 3) Explanatory machine learning based classification models to understand the factors influencing terminal predictions. 4) Model-induced cost analysis to capture the monetary value of the prediction model including assessing the cost implications of misclassifications. An empirical study conducted at a seaport shows that our framework yields cost savings ranging from 14.90% to 30.45% compared to the Business-as-Usual scenario. Incorporating unstructured data as an additional feature in the machine learning models improves prediction performance by up to 6%. Moreover, integrating this framework into the existing operational system poses minimal risk and can be seamlessly executed. Additionally, the proposed methodological framework and its four components has broad applications beyond the shipping industry. | 1. Introduction 1.1. Motivation Container shipping carries over 50 % of world seaborne trade by value (Lee and Song 2017) with container ports serving as pivotal connectors between seaborne and inland transport networks. However, increasing trade volume, disruptive events, and coordination challenges among stakeholders frequently lead to port congestions, disrupting global supply chains (Watkins, 2021). To address this * Corresponding author. E-mail addresses: [Ying.Xie@cranfield.ac.uk](Ying.Xie@cranfield.ac.uk) (Y. Xie), [Dongping.song@liverpool.ac.uk](Dongping.song@liverpool.ac.uk) (D.-P. Song), [Jingxin.Dong@newcastle.ac.uk](Jingxin.Dong@newcastle.ac.uk) (J. Dong), [Yuanjun.Feng2@liverpool.ac.uk](Yuanjun.Feng2@liverpool.ac.uk) (Y. Feng). [https://doi.org/10.1016/j.tre.2025.104331](https://doi.org/10.1016/j.tre.2025.104331) Received 28 September 2024; Received in revised form 19 February 2025; Accepted 21 June 2025 1366-5545/© 2025 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license ([http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/licenses/by/4.0/)). issue, several strategies can be employed including expanding port capacity, enhancing supply chain coordination and operational efficiency. From the perspective of port or terminal operators, one of the most readily available approaches is to improve the operational efficiency, especially in yard management (Hu et al. 2021), quayside operations (Al

FAQs

Why predict out-terminals for imported containers at seaports?

To mitigate yard density and port congestion by enabling more effective container storage decisions after containers are discharged from vessels.

What are the four key components of the proposed framework?

Integrating structured and unstructured data, using practice/knowledge-informed feature engineering, applying explanatory machine-learning classification models, and performing model-induced cost analysis including misclassification costs.

How does incorporating unstructured data affect prediction performance and outcomes?

Adding unstructured data as an additional feature improves prediction performance by up to 6%, and the framework achieves cost savings of 14.90% to 30.45% versus a Business-as-Usual scenario.

Aug 4, 2026

Upload by Quinn